Alkermes plc (ALKS) Earnings Call Transcript & Summary

Okay. Good morning. Thanks, everyone, for joining us at the 35th Annual Piper Sandler Healthcare Conference. I'm David Amsellem from the pharma team. And with us is Alkermes and joining us the CEO, Richard Pops. Delighted to have you with us. And lots to talk about. So I know everyone wants to talk about our favorite topic, orexin agonist 2680. I am not going to start with that; however, we'll start with a high-level question, if that's okay. Since you completed the separation of the oncology business.

I'm particularly interested in now, how are you thinking about the neuroscience business, the stand-alone Neuroscience business, particularly the cost structure, your margins and, ultimately, your strategic vision for that business over the long term.

Well, first of all, it's great to see you, and thanks for having us. I feel like I've been to 35 of them. But it's a cool moment because as simple as it sounds, separating the oncology bit from Alkermes, is going to reveal when we particularly -- more explicitly when we guided in 2024 of what the stand-alone neuroscience company actually looks like. And how few comparables there are to this. This is going to be a profitable company to $1 billion plus top line growing products and a pipeline. And that pipeline will continue to expand, but the identified parts of the pipeline, 2680 and its derivatives are really exciting areas of neuroscience drug development with a different risk profile than what we've historically done in psychiatry. So we see this business as being a profitable business. We've established these profitability targets simply to put markers in the ground to make explicit the fact that we see it as a profitable company. It's going to be cash generative. And I think it provides an amazing platform for continuing to build a neuroscience franchise.

Okay. So I wanted to get your thoughts on M&A and in-licensing. I know this is something that you've talked about in the past. But I wanted to get your latest thoughts on key strategic priorities. Is it bolstering the pipeline by acquisitions or in-licensing? Is it bringing in an asset where you can leverage the commercial infrastructure you have in place? Is it may be both? How are you thinking about that these days?

Well, I think that underscores the differentiated nature of this platform that Alkermes will be because there's a scientific capability that we've honed over many years. We've gotten many drugs approved. We know how to develop drugs. But also, we've built this commercial infrastructure that's quite distinctive. Commercial infrastructure that's not just the ability to sell pharmaceutical products, the ability to sell pharmaceutical products in complex reimbursement environments at lower price points, higher gross to nets with government overlays, payer overlays, really challenging commercial settings. So to answer the question from -- as you think about the stand-alone business, how do you grow that business? Ideally, it would be wonderful to put additional commercial products into that infrastructure because we think we have a competitive advantage. If you're a small company with a single asset, building that infrastructure is perhaps cost prohibitive. And many of the large players aren't in these markets. So we sit with a very few -- a number of other companies as a potential collaborator with those types of assets. On the R&D side, no major stand-alone biopharmaceutical companies made it from internally generated pipeline only. And when capital markets are difficult and cost of capital is high and options for smaller companies are limited, that's when -- you don't have to buy a whole companies also. There's licensing deals with milestones that are attractive for both parties that we'll be looking at as well.

When you think about transactions focused on the pipeline, how are you thinking about R&D risk? I mean do you want to replenish the early-stage pipeline? Do you want to add something that's late stage? And this sort of gets into how big of a deal you'd contemplate, but just help us understand philosophically what you're thinking about?

So think about it with 2 different parameters. One is R&D in a bio-pharmaceutical company is essentially a continuous process. So in order to have a robust ongoing R&D enterprise, you have things sequenced all through early delay -- early being relatively inexpensive and higher risk, later as you cone down and discount rates come down, you could allocate more capital. But the other thing is operating within the constraints of these profitability measures that we've also established because that actually is a good thing because it bounds your R&D spend. It says, okay, we have to allocate capital competitively across different programs. So the fittest survive. And I can tell you, having often when you build a company from scratch when you're living completely on equity funding where you spend every dollar that you raise, you don't have a commercial business. You really -- you're limited only by the amount of money you can raise. When your commercial ongoing enterprise funding yourself, that capital allocation strategy becomes actually quite strategic and I think it's really healthy in terms of R&D allocation. So we'll expand the pipeline in earlier and mid-stage things most likely, but we'll do so within the constraints of the profitability targets that we're going to maintain.

Okay. That's very helpful. So now we can talk about all things orexin...

It took a while. 6 minutes.

6 minutes, right? I mean, so that's a record these days. So I'll start with some news flow actually. So my understanding is that your -- one of your potential competitors, Jazz this morning, talked about pausing their orexin agonist for safety reasons and now evaluating a backup compound. So that's JZP441. So this sort of leads me to a question about how these molecules behave. We saw Takeda have to discontinue one of their orexins due to liver tox, and we have Jazz citing, I believe it was cardiovascular issues. So just talk philosophically about this class and what appears to be a wide variability in terms of how these agents are behaving from a safety perspective.

I smile because there's been so much investor interest in preclinical and Phase I compounds. It should remind everybody that small molecule drug development is inherently very risky. And drugs fail for all kinds of reasons. Some of them are quite dramatic and some are quite prosaic. So our thesis from the beginning. And the reason we got into the chemistry with this was that we realized that the number of parameters you would need to optimize to have a real viable orexin 2 receptor agonist were multiple. And the molecules that were in development were likely to be very different because of the various optimization parameters that need to be addressed. So just a brief litany of that. You have to have very potent orexin 2 specific agonists, okay, as small molecules. Pause for a second there. This is a G-protein-coupled receptor in the brain. You need a small molecule agonist that by itself is challenging, that's to be orally bioavailable challenge. It needs to cross the blood brain barrier challenging. It doesn't -- it can't be a substrate to get pumped out of the blood brain barrier out of the brain, challenging. And then it has to have a pharmacokinetic wave form that's consistent with the sleep wake cycle. So if you do all that and the drug only lasts for 2 hours in the CNS, you've got to give it 5 times a day. Or if it lasts too long and people stay up all night, you don't have a drug. So the probability that all these various drugs are going to look the same in the clinic, we established or we assumed it was going to be very low. So I'm not surprised to see them differentiating. So far, ours is hitting the parameters that we designed it to do. But I think that at the end of the day, not everybody is going to make it to the finish line, the ones that do will be different from each other.

And I want to come back to safety and tolerability, but just wanted to ask you specifically about 2680. So you're moving into a Phase II. Can you talk about the contours of the Phase II study in narcolepsy type 1 dosing treatment duration, even how many sites. And maybe also talk us through how you're thinking how long it will take to enroll such a study?

So let me contrast it first to what we've done historically in psychiatry, where you see a signal of Phase I that might be elaborated in Phase II as you go to Phase III, the studies get very big, very expensive and even more risky because there's higher placebo response. There's more heterogeneous patient populations. It's a white-knuckle ride until you unblind it and hope that you have an alpha less than 0.05. This is the exact opposite in many ways. The phenotype, the disease, narcolepsy type 1, is a deficiency of orexin. In our first cohort of patients essentially -- literally, all the patients respond and they respond very similarly. So the statistical powering is quite straightforward. We showed p-values with 4 patients. So the sample size is not driven by some response rate that you're hoping to define. It's really driven more by safety exposures across the whole program. So a Phase II program, a very simple design. Now that we've honed in on NT1 what the doses would be, 3 doses likely versus placebo, parallel design, maintenance and wakefulness test over 4 to 6 weeks and with an extension thereafter for safety. So the pairwise comparison versus placebo will be on MWT. We've already shown that we're going to affect that. And so then you're really building the safety database throughout the Phase II, Phase III program to satisfy regulatory authorities for registration.

And do you have a sense of how many sites and geography?

We're figuring that out right now. It will be global, for sure. A number of sites TBD, but the cohort sizes typically tend to be 20 to 30 patients per ARM.

Okay. And you're going to look at that [ cohort ] as a secondary measure?

We'll look at -- so the primary endpoint will be MWT. Cataplexy and NT1 is going to be important. And then we'll look at various quality of life, patient reported outcome measures in Phase II to figure out what the best one to capture for Phase III would be.

Okay. That's helpful. So I wanted to come back to tolerability. One of the things, and I don't want to spend too much time on the Jazz compound. But one of the things they did cite, my understanding, is visual disturbances. And we've seen that. So how do you think about the incidence of visual disturbance with your molecule and the extent to which that's just a byproduct of its wakefulness properties.

I think it will be more fully elaborated in Phase II. I'll be curious to see what the Jazz data are with respect to that.

And that was in sleep-deprived healthy...

So we saw very mild transient visual disturbances in a few patients, in the healthy volunteers, not sleep-deprived, but healthy volunteers. So these are people with an intact orexin hypocretin circuitry, albeit only in 4 patients that -- we have not seen it yet in the patients, although in Phase II, we'll be paying attention to that. And now with another observation from Jazz. Now back to the original point, though, about the differences in the molecules, potency, we think matters a lot in this, both with respect to cardiovascular effects as well as other on or off-target effects. So for example, you would have thought with certain earlier generation orexin agonist that you might see blood pressure and heart rate moving concomitantly with your wakefulness. With our compound, separated dramatically, we did not hit a maximum tolerated dose in the healthy volunteer study, yet we see potencies on wakefulness between 1 and 8 milligrams without any effect on cardiovascular. So I think the drugs are going to separate based on their potency and their pKa as well.

Another AE that I thought was noteworthy, not troubling, I should say, but just noteworthy is daytime urinary urgency. This is something we saw with the Takeda compound. We saw it to some extent with yours. How do you think about whether that's an on-target or off-target effect and maybe help us understand the, I guess, for lack of a better term, the extent to which it's troublesome.

It's definitely on target. We think that's absolutely on-target. It's called pollakiuria, if you see it in the tables. And it's not necessarily a frequency of urination, but the impulse, the desire to. And what will be interesting to see is, over time, we've heard some reports that, that diminishes over time as patients get adapted to that sensation and it's probably dose-dependent as well. So we'll be looking at that in Phase II, but I don't expect it to disappear. I think it's very much an on-target phenomenon.

So let's talk about efficacy in the wakefulness signal that you saw in the number of NT1, I think the 4 patients with NT1. How do you frame that signal relative to other daytime wakefulness agents, thinking about products like pitolisant and solriamfetol. What's your general view? And I know it's a small end, but based on what we've seen, how do you think about positioning relative to these other wakefulness products?

Well, I think this is why there's so much excitement in the field because our results build on previous clinical results from Takeda that suggest that it's really quite a dramatic difference from what proceeds. So the two ways, as you know, of treating the disease now are -- one is focused on the sleep side, which is by using oxybates to sleep better. And then the other on the wakefulness side has been drugs like modafinil and other stimulants. So in the classic measure, this maintenance of wakefulness test, where the sleep latency, patients with NT1 will fall asleep in this experience where you put people in a dark room, low light, picked up in EEG, people with NT1 will fall sleep in our case, 3 minutes. With those agents, that 3-minute latency goes into the low teens or high single digits. So you gain a few minutes of latency. It's a 40-minute test and what we and Takeda have shown with orexin, I guess, is that you can max out the test. And not only that, but what we showed with our first 4 patients, we ran that test every 2 days -- every 2 hours on the dosing day. The classic measuring is over 8 hours. We ran it for 10 hours. We maxed it out at 10 hours. And then after we stopped recording, the principal AE that we reported 8 milligrams was insomniac because some of those patients stayed up all night. So really, what we're showing is we can target this wakefulness circuitry in the brain with a very highly potent molecule and within, in dose-dependent way, modulate the duration of that effect.

Okay. You have an NT2 cohort, you have an IH cohort. So 2 questions here. One is when should we expect to see data in those cohorts? And then secondly, just remind us how to think about dosing in those cohorts relative to how you dosed NT1 patients.

We're enrolling NT2 and IH patients now in that same experimental design where we'll do cohorts of 8. And we'll probably analyze the NT2 as the full 8 because we expect more variability in the NT2s. IH is a fairly consistent phenotype. NT2s have variable responses we believe to an orexin agonist. So the hypothesis is that the dosing would need to be 2 to 3x higher, that's based on earlier work in the clinic by others, but also just extrapolating based on our compound. And so I hope that we'll get data from the IH and NT2 cohorts so we can report topline to you guys in the Q1.

Okay. And in IH, you have talked about advancing a different orexin-2 receptor agonist for that population. So with that in mind, any thoughts on when that could go into the clinic? Or is this something where you need to see what the data looks like in IH and then make a go/no-go decision regarding development there?

I think the way that we think about it now is that the differential diagnosis between IH and NT2 is pretty fluid. So if we have a registered drug for NT1, NT2, we will pick up any IH patients. Their diagnosis will change. So if you have a -- if you stipulate that you have another orexin agonist, it's an attractive drug. I don't think you developed it for IH. I think you developed for some other indications that we have in mind that also are characterized by a deficiency in this wakefulness circuitry in the brain.

Okay. So let's move on to LYBALVI. We've got about 6 minutes left, so I want to make sure we...

Just on that front. There are -- some of those indications could be as bigger or bigger than the narcolepsy indication. And you think of it, they would be at a different price point, but they could be much more generalizable phenomenon. If the circuitry is relevant to patients that don't have an orexin deficiency in their brain. And that's why even in our single ascending dose study, we were looking at EEG, quantitative EEG in healthy and we saw changes in wakefulness bands that were consistent with activating that circuitry in healthy. That's provocative and we're going to build on that in the lab.

So if I'm hearing you correctly, and I think I've heard you talk about this before, but essentially another -- other settings where hypersomnolence as a feature wouldn't necessarily be -- probably wouldn't be an orphan disease setting, would be sort of a much bigger market where the pricing would be non-orphan like, right?

Correct. And you might want different PK, you might want different potency. There could be PRN medicines in that setting. There are a lot of different things to think about. Whereas think of NT1 as sort of the enriched environment, show this. That's a deficiency of orexin, you replace orexin. It's an orphan indication, high price point, high value for the patients, chronic therapy.

Okay. Okay. Now LYBALVI. I want to make sure we spend some time on that. So you do have a DTC campaign here that you kicked off earlier this year. Can you just talk about how that's going to evolve in '24. What I guess I'm getting at without asking you a question about guidance, maybe a cute way of asking that. But how should we think about the level of DTC spend or expansion of DTC efforts in '24 versus what you've been doing for LYBALVI this year?

This is a category, the atypical antipsychotics, particularly focused in bipolar. There's a lot of data. There's a lot of predecessor drugs in commercializing the space. So there's -- analytics are pretty good. DTC becomes an ongoing part of your marketing mix. And the same way you have a field force, you have DTC, you have broadcast DTC, you have digital DTC. So what you try to do is tune it and optimize it each year to get the biggest ROI. We're early in it right now. We can't calculate ROI right now because we're really just getting into it. But we'll carry on in 2024, for sure. And when we guide, we'll guide to what we think the overall top line bottom line will be for the drug, but we'll solve for the DTC based on our annual budgeting process.

Okay. The payer landscape, it seems what you've had been saying maybe a year ago regarding contracting has evolved. And I don't want to put words in your mouth, but it seems like the payer landscape has proven to be pretty benign. And...

Don't put those words in my mouth.

Okay. Nothing. Fair enough. But I guess the question here is, how are you thinking about contracting and the extent to which you need to contract more aggressively, particularly as you're focusing on driving more patient activation in the bipolar setting?

Yes, we will have -- we will continue -- we're in ongoing contract negotiations with payers, and those will continue into 2024. Our strategy, as you know, has been -- in 2022, 2023, has been to hold the line on gross demands, endure some amount of suppression in order to preserve the net price because in our modeling, that drives a higher total net revenue over time. But as LYBALVI grows in volume, when it's first introduced to payers, there's zero volume. So they're not losing anything by not contracting, and they say, well, we want use risk rates to put on the formulary. We say no. We begin to see if the drug has a following. As the drug starts getting used, if physicians want to use it and patients want to be on it, they start reimbursing it without getting a rebate. And now every one of those prescriptions is costing them notionally money because they don't have rebate, so it allows you to have a little bit more equilibrium in the negotiation. But because 3 of them control 80% of the market, ultimately, they're going to get a contract from you. But we feel like as we go into 2024, we've been in a much better position now because we basically doubled what we did the first year. It's a real drug. It's going to be around for a long time. We're willing to contract it at fair rates and they want a contract. So we'll guide in 2024, just like we did in 2023, where we think we'll be from gross to net, but we'll negotiate always to maximize the net revenue.

Okay. And then maybe a longer-term question. Do you think the payer landscape will change at all to the extent that KarXT launches and more broadly, the muscarinic agonist come into the market. And I know it's only schizophrenia, so that's not lost on me. But do you think that has any ramifications for your payer strategy in the payer landscape for LYBALVI?

No, this is a category where the payers are in charge. There are so many generic drugs and many of the generic drugs are effective for patients. So they're going to cycle people through generics as much as they can. The activation energy to get somebody from the generics into a branded is really driven by the setting of care, a place where an office can do the paperwork and go through what they need to go through to put somebody on a branded medication. But even when you get on a branded medication, the average length of therapy on the brand medication is so brief, it's a tragic in the way the drug -- the disease is treated, that there's this continuous flux in the marketplace.

But even to the extent there is that flux, I mean, let's say, you have a couple of muscarinic agonists that enter the market, and let's say they're cleaner and there's better persistence, does that at all impact LYBALVI and your ability to drive new starts and grow the patient footprint?

I'm sorry, so your premises that are cleaner and what?

The cleaner and potentially does better patient persistence.

I don't see why that would be the case. The problem is not with the drugs, the problem is with the disease. And so the most effective agent out there is olanzapine. And even with olanzapine, the duration of therapy is brief because people have schizophrenia. And so that's what you're up against. Unlike in narcolepsy, where patients are going to be motivated to take that drug every day, improve their quality of life, schizophrenic patients, unfortunately, after a period of being treated often either feel like I'm feeling better, I don't need my medicines anymore. Or because of the nature of their disease feel that they don't want to take their medicine anymore.

Okay. So lots of changes in terms of new brands, but ultimately, you don't feel any differently about the LYBALVI opportunity?

No, no. I think the LYBALVI occupies a place in the market that's driven by its efficacy and that derives from a long tradition of olanzapine use. So we don't have to establish our bona fides on efficacy. We just have to make sure that people see that as a branded alternative now, we've essentially made olanzapine usable again as a maintenance therapy.

Great. Well, I know we're out of time because that light is blinking red.

Thank you, David.

So with that, I'll leave it there. Thanks, Richard. Thanks to everyone in the audience.